Major TOM Viewer
🌍
83
Quick View of Samples in the MajorTOM-Core Dataset
Mikolaj Czerkawski
mikonvergence
AI & ML interests
None yet
Recent Activity
new activity about 13 hours ago
Major-TOM/MajorTOM-Core-Viewer:fix-l4-gpu-support updated a Space about 16 hours ago
Major-TOM/MajorTOM-Core-Viewer liked a model 20 days ago
earth-compress/georvq-ae-v2Organizations
fix-l4-gpu-support
1
#3 opened 2 days ago
by 
harshinde
updated a
Space about 16 hours ago
liked a
model 20 days ago
liked a
dataset about 1 month ago
updated a
Space about 1 month ago
liked a
dataset 4 months ago
updated a
dataset 6 months ago
updated a
Space 6 months ago
liked a
dataset 6 months ago
updated a
dataset 6 months ago
updated a
Space 6 months ago
published a
dataset 6 months ago
liked a
dataset 6 months ago
upvoted a paper 6 months ago
Add task category and link to EarthEmbeddingExplorer paper
1
#3 opened 7 months ago
by 
nielsr
liked a
Space 6 months ago
posted an update 6 months ago
Post
270
🥕 Introducing BetaEarth - your own Earth embedding emulator [𝐏𝐫𝐞-𝐑𝐞𝐥𝐞𝐚𝐬𝐞]
The past year has brought many notable embedding products, like AlphaEarth, TESSERA or OlmoEarth. We are entering a phase where embeddings begin to act as a substitute for real observation data.
BetaEarth is an attempt to explore how much one can learn from a model based on its embeddings alone, and whether those embeddings can serve as a useful training target for other models. Huge credit to the AlphaEarth team for releasing the embedding archive openly — it's what made this kind of community-built extension possible.
[𝐁𝐞𝐭𝐚𝐄𝐚𝐫𝐭𝐡 𝐢𝐬 𝐧𝐨𝐭 𝐚 𝐟𝐨𝐮𝐧𝐝𝐚𝐭𝐢𝐨𝐧 𝐦𝐨𝐝𝐞𝐥 𝐛𝐮𝐭 𝐢𝐭 𝐭𝐫𝐢𝐞𝐬 𝐢𝐭𝐬 𝐛𝐞𝐬𝐭]
BetaEarth is a flexible (and relatively lightweight) emulator of the AlphaEarth annual product. It doesn't reproduce AlphaEarth's exact outputs, nor the product, but it reaches ~0.87 cosine similarity on held-out data and retains 97% of downstream land-cover classification accuracy. It only took 1-2 days to train.
It can encode any combination (including multi-temporal) of:
- Sentinel-2 L1C
- Sentinel-2 L2A
- Sentinel-1 RTC
- COP-DEM 30 product
The model weights are open, just like its training data (built exclusively using Major TOM). The GitHub repository provides a script for automated generation of embeddings across any footprint.
You can also try the workflow over small bounding boxes on the free Hugging Face web app!
⚙️ GitHub: https://github.com/asterisk-labs/beta-earth
🖥️ Web App: asterisk-labs/betaearth
🏭 Models: https://huggingface.co/collections/asterisk-labs/beta-earth
🟨 Colab: https://colab.research.google.com/github/asterisk-labs/beta-earth/blob/main/examples/generate_demo.ipynb
🗞️ Pre-print: https://github.com/asterisk-labs/beta-earth/blob/main/docs/beta_earth_preprint.pdf
The past year has brought many notable embedding products, like AlphaEarth, TESSERA or OlmoEarth. We are entering a phase where embeddings begin to act as a substitute for real observation data.
BetaEarth is an attempt to explore how much one can learn from a model based on its embeddings alone, and whether those embeddings can serve as a useful training target for other models. Huge credit to the AlphaEarth team for releasing the embedding archive openly — it's what made this kind of community-built extension possible.
[𝐁𝐞𝐭𝐚𝐄𝐚𝐫𝐭𝐡 𝐢𝐬 𝐧𝐨𝐭 𝐚 𝐟𝐨𝐮𝐧𝐝𝐚𝐭𝐢𝐨𝐧 𝐦𝐨𝐝𝐞𝐥 𝐛𝐮𝐭 𝐢𝐭 𝐭𝐫𝐢𝐞𝐬 𝐢𝐭𝐬 𝐛𝐞𝐬𝐭]
BetaEarth is a flexible (and relatively lightweight) emulator of the AlphaEarth annual product. It doesn't reproduce AlphaEarth's exact outputs, nor the product, but it reaches ~0.87 cosine similarity on held-out data and retains 97% of downstream land-cover classification accuracy. It only took 1-2 days to train.
It can encode any combination (including multi-temporal) of:
- Sentinel-2 L1C
- Sentinel-2 L2A
- Sentinel-1 RTC
- COP-DEM 30 product
The model weights are open, just like its training data (built exclusively using Major TOM). The GitHub repository provides a script for automated generation of embeddings across any footprint.
You can also try the workflow over small bounding boxes on the free Hugging Face web app!
⚙️ GitHub: https://github.com/asterisk-labs/beta-earth
🖥️ Web App: asterisk-labs/betaearth
🏭 Models: https://huggingface.co/collections/asterisk-labs/beta-earth
🟨 Colab: https://colab.research.google.com/github/asterisk-labs/beta-earth/blob/main/examples/generate_demo.ipynb
🗞️ Pre-print: https://github.com/asterisk-labs/beta-earth/blob/main/docs/beta_earth_preprint.pdf
updated a
Space 7 months ago
updated a
collection 7 months ago
updated a
Space 7 months ago